Insurers say AI will amplify, not replace, human judgment

AI is becoming a core system question, not just a model question

The most useful takeaway from Sollers Consulting’s CEO Voices Report 2026 is simple: insurers are not describing AI as a replacement plan. They are describing it as a way to make human judgment faster, sharper, and more available where it still matters most, especially in claims, underwriting, and customer service.

That matters because the conversation has moved past novelty. Insurance leaders now identify artificial intelligence as the most influential technology shaping the industry today, ahead of cloud, core systems, and the Internet of Things. The message from this group of 11 CEOs is not that AI will sit on the edge of the business as an experiment. It is already moving into day-to-day operational use across underwriting, pricing, claims handling, customer service, and risk management.

The real shift is from automation theater to operational design

The report’s strongest point is not that AI can automate more tasks. It is that insurers want tools that preserve the parts of the job that still require a human being who can interpret nuance, manage exceptions, and absorb accountability. Empathy, strategic thinking, and personal interaction are becoming more important, not less, because the industry still lives and dies on trust.

That is a direct challenge to software vendors that oversell straight-through processing as the whole answer. In P&C, the hard part is not just generating a recommendation. It is making sure an underwriter can see why the recommendation was made, a claims handler can override it when the file is messy, and a service rep can hand off a delicate conversation without losing context. If the software does not support that flow, it is not ready for real insurance operations.

The report also gives this debate an unusually broad international base. The participating executives represent insurers and organizations across Germany, France, the United Kingdom, Australia, the United States, and Canada, with Europe, North America, and Asia-Pacific all in the mix. That matters because the pressure points are not confined to one market. Carriers everywhere are trying to modernize without breaking local operating models, regulatory expectations, or long-built customer relationships.

What insurers want AI to do in claims and underwriting

The clearest practical reading of the report is that AI should amplify the work of adjusters, underwriters, and service teams rather than hollow it out. In claims, that means faster intake, better sorting of unstructured documents, and cleaner triage, but not fully automated handling of every file. In underwriting, it means better analysis of submission data, stronger risk selection, and more consistent pricing support, while keeping final judgment with the person who understands the portfolio and the broker relationship.

Michał Trochimczuk said AI is already improving efficiency, especially in processing unstructured data. That is the kind of use case insurers understand immediately, because the industry is full of emails, PDFs, photos, notes, and fragmented submission packages that do not fit neatly into legacy systems. He also said cost control will be a key driver of competitiveness, and pointed to automation, simplified standard architectures, modern rating systems, and intelligent risk selection as important investments.

That combination is telling. Insurers are not looking for AI to exist in isolation. They want it wired into the architecture that actually runs the business. If the rating engine is still clunky, if the architecture is too custom, or if the workflow cannot move seamlessly between machine suggestion and human approval, the value gets lost fast.

The buyer checklist is changing

For software buyers and transformation leaders, the report points to a different way of judging vendors. The old question was whether a platform could automate a step in the process. The better question now is whether the platform improves decision quality without degrading trust.

The capabilities that matter most are the ones that preserve human judgment:

• Explainability, so underwriters and claims handlers can see why a recommendation was produced.
• Handoff workflows, so a file can move from automation to human review without losing context.
• Audit trails, so every recommendation, override, and exception is traceable.
• Agent-facing copilots, so service reps get help while still staying in control of the conversation.
• Human-in-the-loop controls, so final decisions on sensitive or ambiguous cases stay with people.

That is the real software design challenge. Insurers do not just need systems that move faster. They need systems that are legible, governable, and usable by frontline staff who are expected to make judgment calls under pressure. In that world, the best product is not the one that hides complexity. It is the one that manages complexity without making the person feel replaced.

Workforce transformation is part of the AI story, not a side effect

Marcin Pluta said several CEOs see workforce transformation as one of the most pressing challenges ahead, and the report rejects the simplistic idea that AI automatically destroys jobs. Instead, roles are evolving toward more technological, analytical, and strategic responsibilities.

That has real consequences for P&C operations. Repetitive administration should shrink, but case management, relationship management, advisory work, negotiation, and portfolio steering become more important. A claims team that used to spend hours gathering documents may spend more time resolving exceptions. An underwriter may spend less time on data entry and more time on portfolio strategy. A service rep may handle fewer rote queries and more sensitive, high-value interactions.

This is where the human side of the report becomes more than a slogan. If AI is going to change the insurer-customer relationship, it has to do so without stripping out the confidence that comes from speaking to someone who can understand the situation and make a call. That is why the report’s emphasis on empathy, personalization, and meaningful customer interaction is so important. It is not sentimental. It is operational.

Why the report carries extra weight

The CEO Voices series was launched on Sollers Consulting’s 25th anniversary, and the company is using it to reflect on 25 years of insurance technology change while looking ahead to the next quarter century. That historical framing gives the report a useful edge. It reminds readers that insurers have already lived through the era of fragmented local systems, weak product control, and slow modernization, and they are now trying to avoid repeating those mistakes in an AI-driven form.

The participant list adds more credibility too. Alongside executives from major insurers, the series includes Marcus Ryu, founder and former CEO of Guidewire, which makes the conversation especially relevant for software and platform buyers. Coverage of the series also points to participants such as Andrew Horton of QBE, a reminder that this is not abstract strategy talk. These are operating leaders thinking about generative AI, cloud computing, and InsurTech innovation in the context of real carriers and real books of business.

The bottom line is clear. AI is no longer being judged only by how much work it can remove. It is being judged by whether it makes the insurer smarter, faster, and more trustworthy without flattening the judgment that defines the business. In P&C, the winning systems will not replace the human. They will make the human harder to replace where it counts.